In comparison with adult onset schizophrenia patients, childhood onset patients have a greater number of pre-psychotic developmental disorders. One fifth of our COS cases have earlier forms of autism spectrum disorder, and one third have motor, speech, language, attention or impulse related developmental disorders (Rapoport et al. 2009; Driver et al. 2013). In contrast, the non-psychotic siblings do not have an elevated rate of early neurodevelopmental disorders. Cognitive studies within this protocol are being completed to obtain prospective measures of working memory. COS probands have significantly decreased hippocampal volume (Mattai et al. 2011). Studies of anatomic hippocampal development in the healthy siblings showed no deviance from that seen for the healthy controls however a study in collaboration with colleagues at the University of Chicago show anterior hippocampal shape deviation in COS, similar to that seen for adult onset subjects, which are also shared, more subtly, by healthy COS siblings (Johnson et al. 2013). We are completing a higher field strength (7T) study of the hippocampus in COS, their non-psychotic siblings and healthy controls. We are finding very prominent reductions in the dentate gyrus which is important in several cognitive functions that are defective in schizophrenia. Similarly, recent analyses also show that COS probands and their siblings show cortical thinning in the siblings that is more marked than for controls and correlaes with the siblings symptoms of schizotypy (Watsky et al. 2016). We are studying the neurocircuitry development and abnormal connectivity in COS patients and their healthy siblings. Such analyses, particularly in healthy siblings may highlight resilience factors or circuits in the brain that are relevant to early psychosis. A task related functional MRI study in collaboration with Dr. Karen Berman, NIMH, is comparing functional response during two working memory tasks for probands, healthy siblings and controls. Very preliminary data suggest that probands and to a lesser extent healthy siblings demonstrate increased recruitment in several brain regions including the right DLPFC and parietal circuits suggesting a subtle functional endophenotype (Loeb et al. in preparation). A recent study explored the small world networks properties using resting state fMRI data in COS patients which suggested loss of shorter connections while the longer connections were relatively preserved (Alexander-Bloch et al. 2014). A resting state fMRI study of COS, their siblings and controls in collaboration with Dr. Stephen Gotts, NIMH, found striking abnormalities in the integration of activation in cognitive/social circuitry and motor/sensory circuits (Berman et al. 2016). Interestingly, these abnormalities bear some similarity to those seen in autism (Gotts et al. in preparation). Several genetic studies are ongoing. There is a higher rate of rare copy number variants (CNVs), with 126 probands, involving 75 proband/sib pairs, and access to the current Database of Genomic Variation of over 26,000 individuals. A total of 13% of COS probands carry a rare, disease related CNV including 5 cases with the 22q11 deletion. This represents a significantly higher incidence compared to that seen for healthy controls and to that for adult onset schizophrenia patients (p<.05) (Ahn et al. 2013). Common variants may also be more salient for COS as indicated by a higher score for COS probands compared with their healthy siblings (Ahn et al. 2016). To better understand the heritability of COS, exomic sequencing of 73 trios is being completed in collaboration with Matthew State, MD, (UCSF). We are also finding an 11% rate of the same rare CNVs for a group of our patients with severe early onset non-schizophrenic disorders indicating the non-specificity of CNV risk. Extramural funding has just been approved for Dr. State and his group to carry out whole genome sequencing for our COS sample. An ongoing collaborative study with Dr. Ricardo Dolmetsch, Novartis Institute, is studying neuronal activation and neurite development from induced pluripotent stem cell (iPSC) derived neurons from our childhood onset schizophrenia patients. Patients with and without rare disease associated copy number variants (CNVs) are included. Contrast groups will include unrelated healthy controls, and healthy and affected family members (some healthy carriers of risk CNVs). Evidence from iPSC-derived neurons from 22q11 subjects indicated reduced L-type voltage gated calcium responses in pyramidal neurons (Psca et al. in preparation). A new study will compare changes in expression for iPSC derived neurons from our best responders and a matched group of poor clozapine responders. Other studies using our samples are finding iPSC derived neurons to show abnormality in resistance to infection (Passeri et al. 2016) or dysregulationof microRNA (Topol et al. 2016). In collaboration with Dr Hakon Hakornason at the Center for Applied Genomics we are studying the broad clinical penetrance for five schizophrenia related CNVs including some found to be higher in COS (Zhou et al. 2016). We have screened a clinical pediatric population of 61,000 children. Preliminary data shows over 80% likelihood of clinical impairment for carriers of 15q13.3 deletion, 22q11.2 duplication and 16p11.2 deletion (Ahn et al. in preparation). Finally, we continue to explore COS patients show a stronger clinical response to clozapine initially and higher long trm compliance with this medication than do later onset patients (Kasoff et al. 2016). These studies are of importance given the dearth of new treatments in child psychiatry (Merikangas et al. 2013; Rapoport. et al. 2013).